Optimized fractional cloudiness determination from five ground-based remote sensing techniques

A 1 year record of fractional cloudiness at 10 min intervals was generated for the Cabauw Experimental Site for Atmospheric Research (CESAR) (51 degrees 58'N, 4 degrees 55'E) using an integrated assessment of five different observational methods. The five methods are based on active as well as passive systems and use either a hemispheric or column remote sensing technique. The 1 year instrumental cloudiness data were compared against a 30 year climatology of Observer data in the vicinity of CESAR (1971-2000). In the intermediate 2-6 octa range, most instruments, but especially the column methods, report lower frequency of occurrence of cloudiness than the absolute minimum values from the 30 year Observer climatology. At night, the Observer records fewer clouds in the 1-2 octa range than during the day, while the instrumental techniques registered more clouds. During daytime the Observer also records much more 7 octa cloudiness than the instruments. A reference algorithm was designed to derive a continuous and optimized record of fractional cloudiness. Output from individual instruments were weighted according to the cloud base height reported at the observation time; the larger the height, the lower the weight. The algorithm was able to provide fractional cloudiness observations every 10 min for 99.92% of the total period of 12 months (15 May 2008 to 14 May 2009).